12. civil - ijcseierd - assessment of spatio-temporal occurence of - tena bekele adgolign

8/9/2019 12. Civil - IJCSEIERD - Assessment of Spatio-Temporal Occurence of - TENA BEKELE ADGOLIGN

1/16

www.tjprc.org [email protected]

ASSESSMENT OF SPATIO-TEMPORAL OCCURRENCE OF WATER

RESOURCES IN DIDESSA SUB-BASIN, WEST ETHIOPIA

TENA BEKELE ADGOLIGN1, G. V. R. SRINIVASA RAO

2 & YERRAMSETTY ABBULU

3

1Research Scholar, Department of Civil Engineering, College of Enginering (A),

Andhra University, Visakhapatnam, Andhra Pradesh, India

2Professor, Department of Civil Engineering, College of Enginering (A), Andhra University,

Visakhapatnam, Andhra Pradesh, India

3Associate Professor, Department of Civil Engineering, College of Enginering (A), Andhra University,

Visakhapatnam, Andhra Pradesh, India

ABSTRACT

The spatial and temporal occurrence of water resources within watersheds should be properly known for effective

resource planning and management. However, direct measurement of parameters describing the physical system is time

consuming, costly, tedious, and often has limited applicability; hence, the use of simulation model is becoming attractive.

The main objective of the study was to assess the spatial and temporal occurrence of surface water resources in Didessa

Sub-basin of the Abbay (Upper Blue Nile) Basin, West Ethiopia. The Soil and Water Assessment Tool (SWAT) was used

to model the hydrology of the sub-basin with dataset including soils, land use/cover, digital elevation model, flow and

meteorological data [from both conventional meteorological stations and the National Centers for EnvironmentalPrediction’s (NCEP) Climate Forecast System Reanalysis (CFSR) grids]. The study showed that monthly and annual

stream flow and hydrologic components in Didessa Sub-basin were predicted by the SWAT hydrologic model with very

good values of model performance evaluation parameters. Besides, the study showed that in data scarce areas, the use of

satellite data from CFSR grids in conjunction with datasets from few conventional weather monitoring stations can give

reliable results of monthly and annual stream flow.

KEYWORDS: SWAT, Hydrologic Modeling, Didessa Sub-Basin, CFSR, Water Resources, Water Yield

INTRODUCTION

Proper planning and management of water resources is vital for wise utilization and sustainable development of

the resource. The total renewable surface water resources of Ethiopia are estimated at 122 BCM (billion cubic meters) per

year from 12 major river basins, and 22 lakes. Renewable groundwater resources are estimated to be about 2.6 BCM

(The World Bank, 2006).

Although there is a universal perception that Ethiopia has adequate water resources, the spatial and temporal

occurrence of these resources within a watershed should be properly known for proper planning and management to be

effective. The use of simulation modeling, which is concerned with the problem of making inferences about physical

systems from measured output variables of the model (e.g, river discharge, sediment concentration), is becoming attractive

because direct measurement of parameters describing the physical system is time consuming, costly, tedious, and often has

International Journal of Civil, Structural,

Environmental and Infrastructure Engineering

Research and Development (IJCSEIERD)

ISSN(P): 2249-6866; ISSN(E): 2249-7978

Vol. 5, Issue 1, Feb 2015, 105-120

© TJPRC Pvt. Ltd.

http://www.tjprc.org/http://www.tjprc.org/http://www.tjprc.org/


2/16

106 Tena Bekele Adgolign, G. V. R. Srinivasa Rao & Yerramsetty Abbulu

Impact Factor (JCC): 5.7179 Index Copernicus Value (ICV): 3.0

limited applicability (Abbaspour, Vejdani, & Haghighat, 2007). The problem with obtaining measured data becomes worse

when the watershed under consideration is very large; hence the use of simulation models becomes mandatory.

The main objective of the study was to assess the spatial and temporal occurrence of water resources in Didessa

Sub-basin of the Abbay Basin, West Ethiopia. The hydrologic components of Didessa Sub-basin were evaluated using the

Soil and Water Assessment Tool (SWAT) model. Besides, the hydrologic prediction efficiency of the Climate Forecast

System Reanalysis (CFSR) dataset in conjunction with dataset from a few conventional meteorological stations in data

scarce areas was also evaluated.

MATERIALS AND METHODS

The Study Area

Didessa Sub-basin forms the southwestern part of Abbay Basin in West Ethiopia. It is situated between 07042’40”

to 09058’17” N latitude and 35

033’14” to 37

00752” E longitude (Figure 1). The total area of the watershed is about 28,092

square kilometres.

The altitude in Didessa Sub-basin ranges between 609 m.a.s.l at the Didessa-Abbay confluence and 3,211 m.a.s.l

at the source of Anger River, one of the major tributaries of Didessa River, in Abe Dongoro District of Horro Guduru

Wollega Administrative Zone, Northeast of the Sub-basin. The highlands in the northeastern and southern parts of the

sub-basin are higher in altitude than 2100 m.a.s.l. The lowlands, with altitudes less than 1,100 m.a.s.l are located at the

eastern remote areas of Anger Sub-watershed and the northern end of the sub-basin, following the valley of Didessa River.

Figure 1: Location Map of Didessa Sub-Basin

Dataset and Data Sources for SWAT Model

The major data types that were used in this study include climate, hydrology, soils, land use/land cover data and

Digital Elevation Model. The climate variables required are daily precipitation, maximum and minimum air temperature,

solar radiation, wind speed and relative humidity. Table 1 summarizes the dataset used in SWAT modeling of the study

area, and their corresponding sources.


3/16

Assessment of Spatio-Temporal Occurrence of Water 107

Resources in Didessa Sub-Basin, West Ethiopia


Table 1: SWAT Input Data Types and their Corresponding Sources

Data

TypeSource Scale/Period Description Remark

Weather

. National Meteorological

Agency of Ethiopia. Climate Forecast

System Reanalysis’

(CFSR) Global Weather

Data for SWAT)

1980-2012

Daily precipitation,

maximum andminimum

temperature, mean

wind speed and

relative humidity

Many of thestation have

long period

missing.

Hydrology

Ministry of Water,

Irrigation and Energy of

Ethiopia

1980-2008Daily and monthly

flow data

Many of the

station have

long periodmissing.

Land

use/cover

Ministry of Water,

Irrigation and Energy of

Ethiopia

1998Land use

classification map

SoilsMinistry of Water,

Irrigation and Energy1998

Soil classification

map

Terrain 20 metersDigital Elevation

Model

Scarcity of weather data is one of the major problems impairing hydrologic modeling and, thereby, proper

planning and management of water resources in data scarce watersheds. To minimize the effects of the problem, alternative

data source is available. For instance, Dile and Srinivasan (2014) evaluated the performance of SWAT model with weather

data from the National Centers for Environmental Prediction’s (NCEP) Climate Forecast System Reanalysis (CFSR)

(Global Weather, 2014). They concluded that in data-scarce regions such as remote parts of the Upper Blue Nile basin,

CFSR weather data could be a valuable option for hydrological predictions where conventional gauges are not available.

Similarly, Fuka, et al . (2013) presented a method for using the CFSR global meteorological dataset to obtain historical

weather data, and demonstrated its application to the modeling of five watersheds representing different hydroclimate

regimes. They reported that utilizing the CFSR precipitation and temperature data to force a watershed model provides

stream discharge simulations that are as good as or even better than models setup using data from traditional weather

gauging stations. Didessa Sub-basin is one of such watersheds with scarce conventional meteorological stations. Hence, in

addition to dataset from four conventional meteorological stations, dataset from three data points of CFSR grids have been

used (table 2). The climate variables include daily minimum and maximum temperature, precipitation, humidity, sunshine

hours and wind speed. Among the four conventional gauging stations, Anger and Bedele have been used as weather

generator stations to fill missing data for the conventional meteorological stations.

Table 2: Location Details of the Weather Monitoring Stations

Weather

Monitoring

Station

Coordinates

RemarkLatitude Longitude

Altitude

(m. a. s. l)

Anger 9027

’00” 36

020

’00

” 1350

Used as weather

generator station

Nekemte 9005

’00” 36

027

’48

” 2080

Arjo 8 45 00” 36 30 00 2565

Bedele 8027

’00” 36

020

’00

” 2011

Used as weather

generator station

P80366 7057

’36” 36

033

’47

” 1552 CFSR’s grid

P95356 9 31 12” 35 37 30 1559 CFSR’s grid P98369 9 49 48” 36 52 30

1787 CFSR’s grid



4/16



Daily and monthly discharge data at different gauging stations in Didessa Sub-basin were obtained from the

Ministry of Water, Irrigation and Energy of Ethiopia. Most of the stations have short records and/or many missing data,

which hinders the use of these stations for model calibration. Hence, a flow monitoring station called Didessa Near Arjo,

with relatively long period of recorded data has been used for model calibration and validation. See table 3 forgeographical location of Didessa near Arjo gauging station.

Table 3: Geographical Parameters of Didessa Near Arjo Gauging Station

Flow Monitoring

Station

CoordinatesCatchment Area (km

2)

Latitude Longitude

Didessa Near Arjo 8041

’0” N 36

025

’0

” E 9,881.0

Source: Ministry of Water, Irrigation and Energy

Soil parameter data such as soil depth, soil texture, hydraulic conductivity and bulk density, and land use data

which are required for the hydrological modeling were obtained from the Ministry of Water, Irrigation and Energy of

Ethiopia. Besides, the recent description of land use by Oromia Water Works Design and Supervision Enterprise

(unpublished report) was adopted. See, respectively, table 4 and Figure 2 for detailed distribution and map of the

sub- basin’s soils.

Table 4: Detailed Distribution of Soil Types in Didessa Sub-Basin

Sr. No. Soil Name Area (ha) Area (%)

1 Dystric Leptosols 15,761.84 56.108

2 Eutric Leptosols 5,336.07 18.995

3 Dystric Cambisols 2 ,088.92 7.436

4 Eutric Regosols 1,767.55 6.292

5 Eutric Vertisols ,447.30 5.1526 Haplic Alisols 1,205.15 4.290

7 Haplic Acrisols 215.18 0.766

8 Haplic Arenosols 159.84 0.569

9 Rhodic Nitisols 101.13 0.360

10 Haplic Nitisols 8.43 0.030

11 Eutric Fluvisols 0.28 0.001

Total 28,092 100

Figure 2: Soil Types in Didessa Sub-Basin


5/16




A Digital Elevation Model (DEM) with 20 m resolution was used to derive the primary terrain attributes of slope

and specific catchment area. See table 5 for detailed slope distribution and Figure 3 for elevation map of the study area.

Table 5: Detailed Slope Distribution in the Watershed

Sr. No. Slope Range (%) Area (%) Area (km2)1 0-3 2.56 719.15

2 3-6 7.44 2,090.04

3 6-12 20.71 5,817.85

4 12-30 46.78 13,141.42

5 Above 30 22.51 6,323.50

Total 100.00 28,092

Figure 3: Elevation Map of Didessa Sub-Basin

Didessa Sub-basin is predominantly covered with woodlands, followed by agricultural practices. The central and

southern parts of the basin are predominantly cultivated (Figure 4).

Figure 4: Major Land Use/Cover Distribution in the Watershed



6/16



Software and Models Used

ArcSWAT, an ArcGIS extension, which is a graphical user interface for the SWAT (Soil and Water Assessment

Tool), was used to model the hydrology of Didessa Sub-basin. SWAT is a river basin, or watershed, scale model developed

to predict the impact of land management practices on water, sediment, and agricultural chemical yields in large, complex

watersheds with varying soils, land use, and management conditions over long periods of time (Winchell, Srinivasan,

Luzio, & Arnold, 2009).

SWAT has been employed to model watersheds of different scales for different purposes. For example, Asres and

Awulachew (2010) applied SWAT to the Gumara watershed of the Abbay Basin to predict sediment yield and runoff, to

establish the spatial distribution of sediment yield and to test the potential of watershed management measures to reduce

sediment loading from hotspots. Similarly, Chekol, et al. (2007) used SWAT model to assess the spatial distribution of

water resources and analyze the impact of different land management practices on hydrologic response and soil erosion in

the Upper Awash River Basin watershed, Ethiopia, so as to gather information for effective watershed management. Bothgroups of investigators reported that they obtained good results.

Likewise, Singh, et al. (2013) applied SWAT to the Tungabhadra catchment in India for the measurement of

stream flow, and reported that they obtained a result in which the observed and simulated data had excellent correlation

during monthly calibration time step. Panhalkar (2014) also used SWAT to estimate the runoff of the Sutluj Basin, India,

and reported a successful performance and applicability of SWAT model.

Model sensitivity analysis, the process of determining the rate of change in model output with respect to changes

in model inputs (parameters), was initially performed using SWAT model’s inbuilt procedures, and the most sensitive

model parameters were identified. After the initial setup, the Sequential Uncertainty Fitting version 2 (SUFI-2) algorithm

of SWAT Calibration and Uncertainty Programs (SWAT-CUP) was applied to re-identify the most sensitive model

parameters and calibrate the model (Abbaspour, 2014). Currently, there is a wide application of SWAT-CUP in calibration

and uncertainty analysis of SWAT models. Gebremicael, et al. (2013) applied SUFI-2 to calibrate SWAT model of the

Upper Blue Nile with the intention of analyzing the trend of runoff and sediment changes, and obtained best fitting model

evaluation parameters. Rathjens and Oppelt (2012) applied the SUFI-2 optimization algorithm of SWAT-CUP on a SWAT

model of a sub-catchment of the River Elbe (Northern Germany) to identify the most sensitive model parameters.

Singh, et al .(2013) applied SUFI-2 technique to evaluate the performance of their SWAT model for the stream

flow measurement of the Tungabhadra catchment in India; and they reported that the obtained results showed good

harmony between the observed and simulated discharge at the 95% level of confidence (95% prediction

uncertainty/95PPU).

SWAT-CUP was setup with the calibration parameters on table 6, and their corresponding maximum and

minimum range of values, which was taken from SWAT Users Guide. SWAT-CUP was iterated and the fitted values for

the most sensitive parameters affecting the flow were identified.


7/16




Table 6: SWAT Calibration Parameters with their Respective Sensitivity Ranking and Fitted Values

Parameter

NameDescription and Units

Sensitivity

Rank

Range of Values Fitted

ValueMinimum Maximum

CN2

Initial Soil Conservation Service (SCS)

runoff curve number for moisturecondition II (-)

1 35 98 55.6

EPCOPlant evaporation compensation factor

(fraction)12 0 1 0.435

ESCOSoil evaporation compensation factor

(fraction)10 0 1 0.689

OV_N Manning’s “n” value for overland flow 13 0.01 30 6.758

ALPHA_BF Base flow alpha factor (decimal) 7 0 1 0.263

REVAPMNThreshold depth of water in the shallow

aquifer for revap to occur (mm)11 0 500 15.5

GWQMNThreshold depth of water in the shallow

aquifer for return flow to occur (mm)3 0 5,000 215.0

SHALLST

Initial depth of water in the shallow

aquifer (mm) 5 0 1,000 321.0

DEEPSTInitial depth of water in the deep aquifer

(mm)8 0 3,000 945.0

GWHT Initial groundwater height (m) 9 0 25 24.525

GW_SPYLDSpecific yield of the shallow aquifer

(m3/m

3)

6 0 0.4 0.025

SOL_AWC Available soil water capacity (mm/mm) 2 0 1 0.999

GW_DELAY Groundwater delay time (days) 4 0 500 19.5

Model Sensitivity Analysis, Calibration and Validation

Model calibration is the process of estimating model parameters by comparing model predictions (output) for a

given set of assumed conditions with observed data for the same conditions, while model validation involves running a

model using input parameters measured or determined during the calibration process (Moriasi, et al , 2007). To perform

such studies as the evaluation of the impact of alternative land management practices on stream water quality and quantity,

first the model must be calibrated and validated for existing conditions (Arnold, et al, 2011). Proper model calibration is

important in hydrologic modeling studies to reduce uncertainty in model simulations (Moriasi et al , 2007). By citing

different sources, Moriasi et. al. (2007) suggests processes that should involve in ideal model calibration as: (1) using data

that includes wet, average, and dry years, (2) using multiple evaluation techniques (3) calibrating all constituents to be

evaluated. The calibration/validation process consists of three steps (Neitsch, Arnold, Kiniry, & Williams, 2011):

Selecting some portion of observed data

Running the model at different values for unknown parameters until fit to observations is good

Applying model with calibrated parameters to remaining observations.

The SWAT model for Didessa Sub-basin hydrology was calibrated for recorded data at Didessa Near Arjo flow

monitoring station. As no automatic calibration procedure can substitute for actual physical knowledge of the watershed,

which can translate into correct parameter ranges for different parts of the watershed (Arnold, et al, 2012), the calibration

procedure involved sensitivity analysis followed by semi-automated calibration procedure by SWAT-CUP, where, at

times, manual manipulation on the selection of calibration parameters was necessary.



8/16



Didessa Near Arjo station was selected for the mere reason that it has relatively better data availability (long term

recorded data), although the amount of missing data is still large.

The station is an outlet for 9,711 km2 (35%) of the total basin area. The years 1980 and 1981 were used as a

warm-up period to enable the model to initialize smoothly (Abbaspour K. C, 2014). Monthly flow data for 1982-1986 and

1992-1996 were used for calibration and validation, respectively. The Nash – Sutcliffe Efficiency (NSE) was used as an

objective function to calibrate and validate the model using 13 flow sensitive input parameters included in the model

(table 6).

The model was calibrated using the Sequential Uncertainty Fitting (SUFI-2) algorithm of SWAT-CUP,

an interface that was developed for SWAT (Abbaspour K. C, 2014). SUFI-2 identifies a range for each parameter in such a

way that upon propagation: 1) the 95% prediction uncertainty (95PPU) between the 2.5th and 97.5th percentiles contains

(brackets) a predefined percentage of the measured data, and 2) the average distance between the 2.5th and 97.5th

prediction percentiles is less than the standard deviation of the measured data (Abbaspour K. C, 2005).

Calibration Criteria

After extensive analysis of published literature on watershed modeling, Moriasi et al, (2007) has developed

general model evaluation guidelines for a monthly time step. Accordingly, they concluded that a general visual agreement

between observed and simulated constituent data would indicate adequate calibration and validation over the range of the

constituent being simulated. They recommended that visual observation should be followed by calculation of values for

three model performance evaluation criteria: Root Mean Square Error (RMSE)-observations standard deviation ratio

(RSR), Nash-Sutcliffe Efficiency (NSE), and Percent bias (PBIAS). Moriasi et al, (2007) describe the parameters as

follows:

RMSE-Observation Standard Deviation Ratio (RSR): RSR is calculated as the ratio of the root mean square

error and standard deviation of measured data, as shown in the following equation:

where is root mean square error, is standard deviation of observed data of the constituent being

evaluated, is the observation for the constituent being evaluated, is the simulated value for the constituent

being evaluated, is the mean of observed data for the constituent being evaluated, and is the total number of

observations. RSR varies from the optimal value of 0, which indicates zero RMSE or residual variation and therefore

perfect model simulation, to a large positive value.

Nash-Sutcliffe Efficiency ( ): The Nash-Sutcliffe Efficiency (NSE) is a normalized statistic that determines

the relative magnitude of the residual variance (“noise”) compared to the measured data variance (“information”).

indicates how well the plot of observed versus simulated data fits the 1:1 line. is computed as shown in the following

equation:


9/16




ranges between −∞ and 1.0 (1 inclusive), with =1being the optimal value. Values between 0.0 and 1.0

are generally viewed as acceptable levels of performance, whereas values


10/16



calibration and validation periods. Figure 5 and Figure 6 are graphical representations of the comparison of observed and

simulated flow values for, respectively, the calibration and validation periods at Didessa Near Arjo flow monitoring

station.

Figure 5: Comparison of Observed and Simulated Monthly Discharge for the Calibration Period (1982-1986)

Figure 6: Comparison of Observed and Simulated Monthly Discharge for the Validation Period (1992-1996)

The two figures portray that there is good harmony between observed and simulated monthly flow for both the

calibration and verification periods, i.e, the graphs for observed data and the corresponding simulated values show similar

patterns. This observation was evidenced using the well known model evaluation parameters, which were used to assess

the performance of the model. Among these, the values for Nash-Sutcliffe Efficiency ( ) and the coefficient of

determination ( were above 0.8 for both the calibration and validation periods. -observation standard deviation

ratio (RSR) was 0.36 and 0.45 for the calibration and validation periods, respectively (table 7).

Table 7: SWAT-CUP Simulation Results for the Calibration and Validation Periods

Model Evaluation

Parameter

Simulation ResultsRemark

Calibration Validation

RSR 0.36 0.45 Very good

NSE 0.87 0.80 Very good

PBIAS (%) -2.3 10.7

Very good for

calibration and good

for validation periods.

0.87 0.80 Very good

P-factor 0.77 0.76

R-factor 0.94 0.75

Overall, the model has demonstrated a very good performance.


11/16


12/16



Figure 7: Simulated Average Basin Values for the Main Stem of Didessa River at its Outlet

The average annual groundwater flow and transmission loss are, respectively, 129.55 mm and 3.08 mm.

The results show that the average runoff coefficient in Didessa Sub-basin is 0.21. The total flow from thesub-basin amounts to about 10.71 BCM per year. This accounts for about 19.54% of the total estimated mean annual flow

from the Abbay Basin (54.8 BMC) as measured at the Sudan boarder (Awulachew, et al , 2007).

Water Budget of the Major Tributaries

Outputs of the SWAT hydrological model for Didessa Sub-basin on the water yield of the major tributaries has

also been evaluated. The four major tributaries of Didessa River are Anger, Dabena, Wama and Upper Didessa.

In terms of size of the catchment area and contribution to the total runoff of Didessa River, Anger Sub-watershed

stands first among the tributary watersheds of the sub-basin. The mean annual rainfall in Anger Sub-watershed is about

1,770 mm, and the mean annual runoff is about 428 mm (higher than the figure reported by Awulachew, et al. (2008).

Hence, the runoff coefficient is 0.24. Anger Sub-watershed contributes an annual total runoff of 3.31 BCM (i.e, about 6%

of the total flow of Abbay River as measured at the Sudan boarder) to Didessa and thereby, Abbay.

The other important sub-watershed in Didessa Sub-basin is Dabena Sub-watershed, with a catchment area of

3,341.2 km2. This watershed receives a mean annual precipitation of 1,886 mm, with the highest evapotranspiration rate of

683 mm per year. The mean annual runoff from the watershed is 366 mm, resulting in a runoff coefficient of 0.19.

Wama is among the top important sub-watersheds in terms of share in Didessa Sub-basin’s annual runoff volume.

It drains a catchment area of about 3,373.4 km2. It receives a mean annual rainfall of 1,934.64 mm (the highest among

Didessa Sub-watersheds) and the mean annual runoff is about 448 mm. Hence, the average runoff coefficient in the

watershed is 0.23.

The mean annual rainfall in Upper Didessa Sub-watershed, the reach of Didessa River upstream of its confluence

with Wama River, is about 1,797, and the mean annual runoff is about 358 mm, resulting in a coefficient of runoff of 0.20.

It drains an area of 5,539.60 km2. It is an important sub-watershed where the dam and reservoir for Arjo-Didessa Irrigation

Development Project, a project which is intended to develop about 80,000 ha of land through irrigation, is being

constructed. The results are summarized on table 9 and Figure 8.


13/16




Table 9: Mean Annual Water Budget of the Major Tributaries

No.Name of

Tributary/Reach

Precipitation

(mm)

Mean

Annual Flow

(m3/s)

SURQ GWQ PET ET

Total Annual

Water Yield

Volume

(10

9

m

3

)

Depth

(mm)1 Anger 1,769.97 104.90 87.29 133.3 1,169.4 499.2 3.310 427.78

2 Dabena 1,886.12 38.70 37.46 85.2 1,329.5 683.1 1.221 365.53

3 Wama 1,934.64 47.88 69.05 195.9 1,129.8 524.4 1.511 447.90

4 Upper Didessa 1,796.80 62.87 38.52 144.0 1,188.9 571..9 1.984 358.16

The dam for Arjo-Didessa Irrigation Development Project is being constructed at about 1.5 km upstream of the

confluence of Wama and Upper Didessa rivers. The dam site drains a catchment area of about 5,536.0 km2. The mean

annual flow at the dam site is about 62.82 m3/s (total of 1.983 BCM), a sub-watershed average of about 358 mm.

It can be observed that there is variability in water yield between the tributary watersheds and the basin average.

This may be due to the differences in catchment behavior (land use/cover, soil types and relief).

Figure 8: Mean Monthly Values for the Major Sub-Watersheds of Didessa River

SUMMARY

The SWAT hydrologic modeling of Didessa Sub-basin involved setting up the model using a DEM of 20 meters

resolution, land use/cover, and soils data of the study area. Weather data from four conventional meteorological stations

were used. These stations are concentrating around the middle of the watershed, and hence it was envisaged that the model

setup using dataset from the four stations alone may not predict the intended parameters with sufficient accuracy. Hence,

additional dataset from three grids of the Climate Forecast System Reanalysis (CFSR) were used, making the distribution

of weather monitoring stations fairly uniform across the sub-basin. The model setup was followed by a sensitivity analysis

using SWAT’s in-built default parameters for flow, and the most sensitive model parameters were identified. The final

model calibration and validation were performed using the Sequential Uncertainty Fitting (SUFI-2) algorithm of SWAT

Calibration and Uncertainty Programs (SWAT-CUP).



14/16



The model predicted monthly discharge with high accuracy, with values of 0.87 and 0.80 for the calibration

(1982-1986) and validation (1992-1996) periods, respectively; and values of 0.87 and 0.80 for the calibration and

validation periods, respectively. The calibration and validation of the model were performed for a measured flow data at

Didessa near Arjo gauging station, which is located at 80

41’

0” N latitude and 360

25’

0”

E longitude.

The mean annual precipitation in Didessa Sub-basin is about 1,774.20 mm. The mean annual water yield is about

381.4 mm (10.71 BCM). The water yield is majorly constituted by lateral flow (51%) followed by groundwater flow

(34%). The rest (16%) comes from surface runoff, while about 1% of the flow is lost through transmission. Temporally, the

months of June to September are the period during which the sub-basin receives most of the precipitation. About 71% of

the rainfall comes during this period. Likewise, about 78% of the total water yield occurs during this period. The period is,

of course, the main rainy season in the area. The months of November to February are the period during which the area

receives the minimum amount of rainfall (about 4%). About 4.15% of the annual water yield of the sub-basin is generated

during this period.

Spatially, there is variability in both the amount of precipitation the area receives and the runoff generated,

between the sub-basin average and that of the major tributaries. Draining the widest area of all the major tributaries, Anger

Sub-watershed receives a mean annual rainfall of about 1,770 mm, and generates a mean annual water yield of about 428

mm (3.31 BCM). Wama Sub-watershed receives the highest annual rainfall (about 1,935 mm) followed by Dabena

(1,886 mm) and Upper Didessa Sub-watershed (1,797 mm). Wama Sub-watershed generates the highest rate of runoff

(about 448 mm) followed by Anger Sub-watershed (428 mm) and Upper Didessa Sub-watershed (358 mm). The amount of

water yield is not proportional to the amount of rainfall received, as this should be a function of the watershed

characteristics (land use/cover, soils, and land slope).

CONCLUSIONS

The monthly and annual stream flow of the main stem of Didessa River and its major tributaries has been

predicted by a SWAT hydrologic model, a model which performed very well. Besides, the monthly and annual

precipitation and evapotranspiration have been predicted. Although the values of the parameters which have been predicted

by SWAT in this study are somewhat different from those reported by previous investigators, the values of the statistical

parameters which are often employed to test the predicting efficiency of SWAT model clearly portray that the results of the

present study are quite trustworthy.

The study also showed that in data scarce areas, the use of satellite data from Climate Forecast System Reanalysis

(CFSR) grids in conjunction with datasets from a few conventional weather monitoring stations can give reliable results of

monthly and annual stream flow and other weather parameters. Thus, even though dataset from conventional

meteorological stations are still more desirable for greater accuracy and reliability, data scarcity may no longer hinder

watershed hydrologic modeling exercises in such areas as Didessa Sub-basin.

REFERENCES

1. Abbaspour, K. C. (2005). Calibration of Hydrologic Models: When is a Model Calibrated? Modelling and

Simulation Society of Australia and New Zealand , 2449-2455.

2. Abbaspour, K. C. (2014). SWAT-CUP 2012: SWAT Calibration and Uncertainity Programs-A User Manual.


15/16




Eawag: Swiss Federal Institute of Aquatic Science and Technology.

3. Abbaspour, K, Vejdani, M, & Haghighat, S. (2007). SWAT-CUP Calibration and Uncertainty Programs for

SWAT. Modelling and Simulation Society of Australia and New Zealand , 1596-1602.

4. Abbay Basin Authority. (2014). Socioeconomic Profile of Didessa Sub-basin. Bahir Dar: Abbay Basin Authority.

5. Arnold, J. G, Moriasi, D. N, Gassman, P. W, Abbaspour, K. C, White, M. J, Srinivasan, R, et al. (2012). Swat:

Model Use, Calibration and Validation. American Society of Agricultural and Biological Engineers , Vol. 55(4):

1491-1508.

6. Arnold, J, Kiniry, J, Srinivasan, R, Williams, J, Haney, E, & Neitsch, S. (2011). Soil and Water Assessment Tool

Input/Output File Documentation Version 2009. Texas: Texas A&M University System.

7. Asres, M. T, & Awulachew, S. B. (2010). SWAT based runoff and sediment yield modelling: a case study of the

Gumera watershed in the Blue Nile basin. Ecohydrology and Hydrobiology , 2-4: 191-200.

8. Awulachew, S. B, McCartney, M, Steenhuis, T. S, & Ahmed, A. A. (2008). A Review of Hydrology, Sediment

and Water Resource Use in the Blue Nile Basin. IWMI Working Paper 131 . Colombo, Sri Lanka: International

Water Management Institute.

9. Awulachew, S. B, Yilma, A. D, Loulseged, M, Loiskandl, W, Ayana, M, & Alamirew, T. (2007). Water

Resources and Irrigation Development in Ethiopia. IWMI Working Paper 123 . Colombo, Sri Lanka: International

Water Management Institute.

10. Chekol, D. A, Tischbein, B, Eggers, H, & Vlek, P. (2007). Application of SWAT for assessment of spatial

distribution of water resources and analyzing impact of different land management practices on soil erosion in

Upper Awash River Basin watershed. Catchment and Lake Research , 110-117.

11. Conway, D. (2000). The Climate and Hydrology of the Upper Blue Nile River. The Geographical Journal ,

1: 49-62.

12. Dile, Y. T, & Srinivasan, R. (2014). Evaluation of CFSR Climate Data for Hydrologic Prediction iIn Data Scarce

Watersheds: An Application in the Blue Nile River Basin. Journal of American Water Resources Association,

1-16.

13.

FAO. (2012). Coping with Water Scarcity: An action framework for agriculture and food security. Rome: Foodand Agricultural Organization of the United Nations.

14. Fuka, D. R, Walter, M. T, MacAlister, C, Degaetano, A. T, Steenhuis, T. S, & Easton, Z. M. (2013). Using the

Climate Forecast System Reanalysis as weather input data for watershed models. Hydrological Processes .

15. Gebremicael, T, Mohamed, Y, Betrie, G, van der Zaag, P, & Teferi, E. (2013). Trend Analysis of runoff and

sediment fluxes in the Upper Blue Nile basin: A combined analysis of statistical tests, physically-based models

and landuse maps. Journal of Hydrology , 482: 57-68.

16. Global Weather. (2014, September 10). Global Weather Data for SWAT . Retrieved from Texas A&M University

Spatial Sciences: globalweather.tamu.edu/home



16/16



17. McCartney, M. P, Alemayehu, T, Easton, Z. M, & Awulachew, S. B. (2012). The Nile River Basin: Water,

Agriculture, Governance and Livelihoods. Abingdon, UK: International Water Management Institute.

18. Ministry of Water Resources. (1998). Abbay River Basin Integrated Development Master Plan. Section

XII-Environmnet . Addis Ababa, Ethiopia: Ministry of Water Resources.

19. Moriasi, D. N, Arnold, J. G, Liew, M. W, Bingner, R. L, Harmel, R. D, & Veith, T. L. (2007). Model Evaluation

Guidelines for Systematic Quantification of Accuracy in Watershed Simulations. Transactions of the ASABE:

American Society of Agricultural and Biological Engineers , 885-900.

20. Moriasi, D. N, Arnold, J. G, Liew, M. W, Bingner, R. L, Harmel, R. D, & Veith, T. L. (2007). Model Evaluation

Guidelines for Systematic Quantification of Accuracy in Watershed Simulations. American Society of

Agricultural and Biological Engineers , 50(3): 885−900.

21. Neitsch, S, Arnold, J, Kiniry, J, & Williams, J. (2011). Soil and Water Assessment Tool Theoritical

Documentation Version 2009. Texas: Texas Water Resources Institute.

22. Oromia Water Works Design and Supervision Enterprise. (2013). Didessa and Dabus Sub-basins Land Use Land

Cover Final Report. Finfinne: Unpublished report.

23. Panhalkar, S. (2014). Hydrological modeling using SWAT model and geoinformatic techniques (article in press).

The Egyptian Journal of Remote Sensing and Space Sciences .

24. Rathjens, H, & Oppelt, N. (2012). SWAT model calibration of a grid-based setup. Advances in Geosciences,

55-61.

25.

Singh, V, Bankar, N, Salunkhe, S. S, Bera, A. K, & Sharma, J. R. (2013). Hydrological stream flow modelling on

Tungabhadra catchment: parameterization and uncertainty analysis using SWAT-CUP. Current Science,

104-9: 1187-1199.

26. The World Bank. (2006). Ethiopia: Managing Water Resources to Maximize Sustainable Growth-AWorld Bank

Water Resources Assistance Strategy for Ethiopia. Washington: The World Bank.

27. Winchell, M, Srinivasan, R, Luzio, M. D, & Arnold, J. (2009, May). ArcSWAT 2.1.5 Interface for SWAT2005.

User's Guide . Texas, U.S.A: Blackland Research Center, Texas Agricultural Experiment Station.

12. civil - ijcseierd - assessment of spatio-temporal occurence of - tena bekele adgolign

Documents